Estimation of the relation between the slow rate of growth which followed the first oil shock and the speed of labour adjustment in Japanese manufacturing has produced some contradictory results. This paper argues that the discrepancy is basically due to different data sets being used; in order then to produce a consistent choice of the data set a criterium is brought forward which emphatizes the relevance of qualitative adjustments; estimation based on related data shows how the adjustment speed has generally increased after the shock, and that a negative relation does exist between an index of performance and the adjustment speed itself.